Hacker News: roger_

New comment by roger_ in "Fixing Tool-Call Accuracy in vLLM"

roger_ — Mon, 26 Oct 2026 15:11:52 +0000

Does this work with Qwen 3.8?

New comment by roger_ in "Turn a WiFi 7 Router into a 320Msps Software Defined Radio Receiver"

roger_ — Tue, 13 Oct 2026 11:26:05 +0000

This should have full TX: https://github.com/cyber-murmel/esp-sdr-zephyr

New comment by roger_ in "Build your own decision model"

roger_ — Sun, 11 Oct 2026 11:28:43 +0000

https://en.wikipedia.org/wiki/Platt_scaling

New comment by roger_ in "Pi 1.0"

roger_ — Thu, 01 Oct 2026 20:13:12 +0000

Paseo is great and has mobile notifications but it’s not as polished as omp web.

New comment by roger_ in "A single function Jev-like wrapper for LLMs, including vision models"

roger_ — Sat, 26 Sep 2026 15:47:14 +0000

I guess “A” by itself would be a seperate token but my point is the model might be trying to say something that begins with that letter rather than actually answering.

You’d need to use an approach that links the output to a closed set.

New comment by roger_ in "A single function Jev-like wrapper for LLMs, including vision models"

roger_ — Sat, 26 Sep 2026 11:36:34 +0000

> Answer with the letter of the best option only [A, B, C]

What if it says "D"? What if it tries to say "Additional details needed"?

(Also no calibration, etc.)

New comment by roger_ in "Homebrew 7.0.0"

roger_ — Tue, 15 Sep 2026 13:34:58 +0000

Does that break anything? Surprised it’s not the default.

New comment by roger_ in "Homebrew 7.0.0"

roger_ — Sun, 13 Sep 2026 10:55:13 +0000

Does installation still require root and a dedicated user account on Linux?

That really put me off.

New comment by roger_ in "Technique for Manipulating Satellite Photos Now Reveals Ancient Images (2025)"

roger_ — Thu, 10 Sep 2026 20:09:11 +0000

Always wanted to implement that on a phone, it would be like a superpower!

New comment by roger_ in "iPhone Duo"

roger_ — Wed, 09 Sep 2026 20:57:36 +0000

My iPad is my primary computer so... me.

New comment by roger_ in "Tiny $70 Xteink X3 e-reader"

roger_ — Mon, 07 Sep 2026 15:18:50 +0000

Thanks!

I’m looking for something that pushes portability without compromising too much core usability.

For example I can live without a touch screen if it means high quality physical buttons.

New comment by roger_ in "Tiny $70 Xteink X3 e-reader"

roger_ — Mon, 07 Sep 2026 14:15:49 +0000

Surprised I never heard of this!

Looking to pick one up, any recommendations on X4 Pro vs Classic?

New comment by roger_ in "Defrag98: Windows 98 Disk Defragmenter Simulator Online"

roger_ — Sat, 29 Aug 2026 23:09:01 +0000

My computer's definitely gonna be faster after this.

New comment by roger_ in "25,000 Lbs. Of Chicken Products Recalled in 5 States: USDA"

roger_ — Fri, 28 Aug 2026 22:06:32 +0000

~5000 chickens if anyone cares :(

New comment by roger_ in "AirLLM 70B inference with single 4GB GPU"

roger_ — Mon, 03 Aug 2026 13:14:44 +0000

Seeing a lot of these “run 1TB models with 1GB RAM” projects recently. Most seem vibe coded and probably won’t be maintained.

Hoping a winner emerges with some real momentum behind it.

New comment by roger_ in "Tech note: making your own V-I plots at home"

roger_ — Sun, 19 Jul 2026 01:04:38 +0000

I’m familiar with curve tracers but hadn’t heard about source measure units.

Don’t see any on eBay for dirt cheap, wonder what author paid (he said a fraction of retail).

New comment by roger_ in "Benchmarking 15 “E-Waste” GPUs with Modern Workloads"

roger_ — Mon, 13 Jul 2026 23:34:19 +0000

Getting (further) into this myself so good timing. Running Qwen 3.6 27B at decent speed on some old cards but going to branch out.

I bough an Octominer for ~$150 which has power and PCIe slots and a basic Celeron and should let me expand to as many GPUs as I want.

I considered the P100s but I think the V100 16GBs are a better deal at $250. The 32GBs are way too much though.

New comment by roger_ in "Benchmarking 15 “E-Waste” GPUs with Modern Workloads"

roger_ — Mon, 13 Jul 2026 23:28:24 +0000

That’s cool but 7 - 12 tps is frustrating for anything interactive.

New comment by roger_ in "Mesh LLM: distributed AI computing on iroh"

roger_ — Sun, 12 Jul 2026 12:59:25 +0000

Does this support Qwen 3.6 (e.g. 27B) and the myriad of llama.cpp options (batch sizes, quantization, etc.)?

I'd love to see some performance data.

New comment by roger_ in "Opinionated and easy Pi.dev configuration"

roger_ — Thu, 09 Jul 2026 18:12:29 +0000

Funny you mention that because the OMP GitHub issues (https://github.com/can1357/oh-my-pi/issues) is almost exclusively attended to by a very efficient bot. I routinely submit issues and they’re replicated and resolved (via PRs) in ~30 mins.

To your question: it’s the difference between one person and an agent working on something vs an experienced dev and the entire community.

Time and tokens too of course.